Dr. Kim shares his concerns with ProtecT trial interpretations

In this video, Isaac Y. Kim, MD, PhD, MBA, discusses 3 disagreements that he has with the conclusions being drawn about the ProtecT trial (NCT02044172)–a 15-year outcome study in prostate cancer–which he outlined in the editorial, “Prostate cancer screening and management: Caution against over-interpreting the results of the latest study, ProtecT.” Kim is the chair of urology at Yale School of Medicine in New Haven, Connecticut.

Video Transcript:

The first is the issue of the effectiveness of PSA-based prostate cancer screening. One of the comments I heard after the study came out was that prostate cancer screening being that aggressive may not be that necessary. Again, this goes back to the same issue that we faced more than a decade ago when our United States Preventive Service Task Force had a similar recommendation at the time. Now afterwards, due to the impact, that decision was pulled back a little bit. But nevertheless, again, we're still having this issue of how aggressive do we need to be with prostate cancer screening? And more specifically, this PSA based screening, how effective is it? The concern that I have in trying to answer that question based on this study is I don't think [this is a] study that's designed to answer that question. I think the effectiveness of prostate cancer screening should not be part of the analysis or part of the conclusion that should be drawn from this ProtecT study.

The main reason for that again, is that at the end, after screening for prostate cancer based on PSA screening, not every patient agreed to the randomization. Initially they had a little more than 82,000 men, and 2600 men were diagnosed with prostate cancer. But at the end, only 1600 men agreed to be randomized. Now, you put yourself in the shoes of a patient who has been just diagnosed with cancer, and you know that one of the arms is going to be not doing anything, so called the active monitoring. If you have a high-risk disease or aggressive disease, would you actually [have] agreed to have that study, where again, your chances of being in a group where you will not be treated is about a third of the time? I would not. And I think therein lies the problem here. Yes, 1600 men, which is again, a lot of men, have agreed to be randomized. My concern is that based on this data, and based on logic that most men who have high-risk disease probably have opted out of the study. And in fact, that's what the cohort shows, because at the end, 77% of men in this trial have a grade group 1 disease. For men who have grade group 1 disease, of course they're going to have an excellent outcome. In fact, most of the men who have grade 1 disease at this point in time, I would recommend surveillance also.

I think to that end, it goes to the next point, that it is no surprise that the overall mortality in this study is so low. At the end, patients who are assigned to the active monitoring group, the prostate cancer specific mortality was 3.1%. In the radical prostatectomy group, 2.2%. And then the radiation group was 2.9%. So again, the prostate cancer specific mortality was excellent in this clinical trial. But that number is deceptive in a sense because, as I alluded to before, 77% of the cohort are men with grade group 1 disease, which has a very favorable prognosis. In extending or making this conclusion, I think we have to be very careful in saying is this is relevant to patients who have a very favorable risk stratification after diagnosis.

And then the third point that I have is that because the prostate cancer specific mortality across the 3 groups are similar, one of the [conclusions was] that in patients with a lethal prostate cancer, that treatment decision doesn't alter the course of the disease. Again, this goes to the whole idea that maybe the prostate cancer fate, or the fatal outcome of the prostate cancer patients on diagnosis is already determined, and any intervention we do may not have an impact in alternating the course of disease, which is very, very important for a clinician like me, because I'm hoping that my treatments will help our patients.

To that, I want to take a look at the data again. I don't think that's there. Because if you look at the overall cohort, it's not that everyone in that group or in the active monitoring group were not treated for cancer, and they were observed 15 years. What has happened is that this is an active monitoring group. So over time, based upon the follow-up information, based on the clinical information, the patient will opt in for our intervention. In fact, that is consistent with what we do for active monitoring/active surveillance right now. So, at the end, in patients who had intervention in the active monitoring group were 60% or 61%. Now, if you are patient, and you're in the active monitoring group, and over time you decide to have an intervention done, isn't that more likely than not that you will be the one who is eventually identified to have more aggressive disease? What I'm saying is this: over time, we're selecting out patients who have aggressive disease in that active monitoring group. Therefore, I think it's reasonable to assume that by end of the 15 years, those patients who have aggressive prostate cancer probably were treated for the most part. I will say majority of them, because I think that would be clear by 15 years, if not all of them.

I don't think that you can make an assumption that by not doing anything, you're not making an impact in the course of disease. But in fact, most of the men actually have some sort of treatment done for disease. I think what should be more appropriate is that the initial treatment decision does not impact [survival], but I don't think you can say that intervening aggressively for this lethal disease does not alter the course. I don't think you can make the conclusion based on the data.

This transcription has been edited for clarity.